In the state of the art, it is customary to monitor yarn quality either right after manufacturing (for example, after the spinning position of a rotor spinning machine) or during a handling step following production (for example, during rewinding from one yarn spool to another). As a rule, physical parameters or measured values dependent on them are monitored in this step. It is known, for example, that yarn thickness or hairiness can be analyzed by evaluating the shadow cast by the yarn on a control surface and using this to draw conclusions about yarn quality.
So reliable statements about yarn quality can be made from the measured values determined, however, a reference value used for comparing the measured values determined is always needed. If the measured values deviate from the reference value within predefined limits, then it is assumed that yarn quality conforms to specifications. Generic yarn monitoring is shown in DE 10 2005 017 606 A1, for example.
The reference value is finally determined at the start of a production process by monitoring the measured variable to be monitored over a certain time period and converting it to a reference value.
However, the disadvantage of the state of the art is that the initial determination of the reference value is time-dependent and contains errors. In particular, the quality of the entire yarn batch deviates from the target when the initial determination of the reference value is performed erroneously for any reason.